PRMT5 is a therapeutic vulnerability in cutaneous T-cell lymphoma

Abstract Background: Cutaneous T-cell Lymphoma (CTCL) is a type of T-cell non-Hodgkin's lymphoma generally localized to the skin. Novel, more selective treatment strategies that target pathways driving CTCL growth and survival remain a significant unmet need. Protein arginine methyltransferase 5 (PRMT5) is a symmetric dimethyl arginine (SDMA) transferase which has been shown to be involved in several T-cell survival pathways, including cytokine responsiveness and T-cell receptor signal transduction. Loss of MTAP, a biomarker of PRMT5 inhibitor sensitivity, has been previously shown to be prevalent in several subtypes of T-cell lymphomas, including CTCL. Here, we report on the activity, function, and therapeutic targeting of PRMT5 using in vitro and in vivo models of CTCL. Methods: All cell lines were validated by STR profiling. Genetic knockdown of PRMT5 was performed using Tet inducible SMART shRNA vectors (Horizon discovery). Pharmacological targeting of PRMT5 in vitro was performed with the PRMT5 selective SAM competitive inhibitor PRT382 with quantification of viable cells by flow cytometry. For in vivo treatment, the PRMT5 selective SAM competitive inhibitor PRT808, was given in a chow-based formulation at 10 mg/kg dosage on a 4-day on, 3-day off schedule. Tumor cells and normal immune cells were quantified by flow cytometry. SDMA protein analysis was performed using the scioSDMA profiling assay (Sciomics, Germany). Protein motif enrichment analysis was carried out with pLogo. Results: Previously published CRISPR genetic screening of T-cell lymphoma cell lines (n=8) revealed that, among the PRMT family, both PRMT1 and PRMT5 are necessary for survival. CTCL cells have a significant increase in PRMT5 and SDMA-modified histone and non-histone proteins compared to resting and activated T-cells. We focused our attention on PRMT5 given the clinical feasibility of targeting PRMT5 in patients. We confirmed PRMT5 dependency in CTCL cell lines (Myla and Hut-78) using a Tet-inducible shRNA system. Treatment with PRT382 induced a dosage and time dependent decrease in SDMA levels across CTCL cell lines, with significant (> 90%) loss of SDMA by day 4, with 4 out of 5 cell lines having IC50 values less than 200 nM on day 6 of treatment. Sensitive cell lines underwent G1 cell cycle arrest and cell death upon PRMT5 inhibition. As expected, loss of MTAP was frequent in T-cell lymphoma lines (6 of 8 cell lines) and all cell lines with loss of MTAP were sensitive to PRMT5 inhibition. We confirmed this activity in vivo in both a human cell lined derived xenograft model of CTCL (Myla) and in a murine immunocompetent model of tumor stage CTCL adoptively transferred from IL-15 transgenic mice. Importantly, in this immunocompetent model, we were able to get significant, selective CTCL suppression, with minimal reduction in normal peripheral T-cells(median survival vehicle vs PRT808: 29 days vs 66 days, p <0.001), confirming the specific activity of PRMT5 inhibition in CTCL cells over normal T-cells. Mechanistically, PRMT5 inhibition was able to lead to a dose dependent downregulation of CD132, the common gamma chain cytokine receptor, in CTCL cells, a mechanism previously shown in PRMT5 knockout mouse models. To uncover the SDMA-modified protein landscape in CTCL cells, we developed a novel immunoassay platform capable of detecting SDMA levels across a panel of 1438 unique targets. Proteins with high SDMA signal detection were associated with significant enrichment of (X)RG motifs (+1G enrichment: p<0.0001), the canonical site of arginine posttranslational modification, confirming the specificity of this platform. Using Myla cells treated with and without PRT382 treatment (200 nM, day 4), we found significant SDMA loss in 360 proteins (adjusted p value <0.05) with numerous proteins involved in pathways highly relevant to CTCL, including the Jak-STAT, TGF-beta, and MMP family members. Validation and functional studies of these proteins are ongoing and will be presented at the meeting.Conclusions: Our data show that blockade of PRMT5 activity leads to significant CTCL suppression in in vitro and in vivo models of CTCL. PRMT5 targets multiple CTCL relevant pro-survival proteins for symmetric arginine dimethylation, thus underscoring its multifaceted oncogenic role in CTCL. Further clinical evaluation of PRMT5 inhibitors as a therapeutic strategy in patients with CTCL is warranted.